United States General Accounting Office
GAO
Testimony
Far Releaseon
Delivery Expected
at 1O:OO
a.m. EST
Thursday,
April 14, 1994
MILITARY SPACE
PROGRAMS
Before the Subcommitteeon Acquisition,
Committee on Armed Services,House
of Representatives
ComprehensiveAnalysis
Neededand Cost Savings
Available
Statementof Louis J. Rodrigues, Director,
SystemsDevelopment and Production Issues,
National Security and International Affairs Division
GAO/T-NSIAD-94-164
-\6\35+
Mr.
Chairman
and Members of the
Subcommittee:
I am pleased to be here today to discuss
two Department
of
Defense (DOD) satellite
programs --the Defense Support Program
(DSP) for early missile
warning,
including
its planned
replacement
called
Alert,
Locate,
and Report Missiles
(ALARM),
and the Milstar
satellite
communications
system.
These
multibillion
dollar
space programs have demonstrated
DOD's
commitment
to effectively
deter the former Soviet Union strategic
threat.
However, U.S. military
needs now emphasize tactical
capabilities
for future
regional
conflicts.
As a result,
DOD is
striving
to modify these programs to better
serve tactical
forces.
RESULTS IN BRIEF
DOD's action
to terminate
the most recent DSP follow-on
program
and begin a new effort
provides
an opportunity
to fully
assess
the needs for early missile
warning
information.
DOD's current
plans to replace
DSP with ALARM will
require
a comprehensive
analysis
that includes
(1) reviewing
and validating
operational
requirements,
(2) selecting
the most cost-effective
alternative
from plausible
candidate
systems,
(3) ensuring
that the system
selected
is affordable,
(4) demonstrating
that an adequate
transition
can be made from DSP to ALARM to avoid coverage gaps,
and (5) evaluating
the launch vehicle
implications
for DSP.
Concerning
the Milstar
program,
DOD could save over $2 billion
if
it determines
that not acquiring
the last two satellites
under
the current
plan would not create an undue operational
risk in
the short term.
Such a decision
would need to be accompanied by
a plan to accelerate
the development
of an enhanced Milstar
that
is smaller
in size,
lighter
in weight,
lower in costs,
and
capable of being launched on a smaller
vehicle
than the Titan
IV.
Accelerating
the enhanced Milstar
effort
may require
some
additional
investment
in the short run.
DSP REPLACEMENT DECISION
REOUIRES A COMPREHENSIVE ANALYSIS
DSP satellites
have been deployed
for over 20 years,
and DOD's
efforts
to replace
DSP with more modern systems have encountered
several
setbacks.
For example,
in the early
19808, the Advanced
Warning System never fully
materialized
because of immature
technology
and high costs.
In the late 198Os, DOD spent about $1
billion
on the Boost Surveillance
and Tracking
System before
discontinuing
research
and development
in 1990.
For fiscal
years
1992 through
1994, the Congress appropriated
$515 million
for the
Follow-on
Early Warning System (FEWS) before DOD terminated
the
program in late 1993 based on affordability
reasons.
In late February
1994, the Air Force selected
ALARM to be DSP's
replacement.
Its plans are to release
a draft
request
for
proposal
in June 1994, initiate
research
and development
in
fiscal
year 1995, have a flight-ready
prototype
satellite
available
by late 1997, and deliver
the first
operational
satellite
in 2004.
ALARM is to be smaller
than DSP with
emphasis on greater
support
to tactical
forces.
Requirements
an
Are Not Yet Established
Since program inception,
DSP was designed toward detecting
strategic
missile
launches
from land and submarines,
as opposed
to shorter
range tactical
missile
launches within
a theater
of
operations.
However, during the Persian Gulf War in 1991, DSP
provided
the primary
tactical
warning of Iraq's
surface-tosurface
Scud missile
launches.
DOD's assessment of DSP's
performance
was that sufficient
warning was provided.to
the
Army's Patriot
missile
defense system, but that an improved
sensor capability
would be needed for the future.l
During 1989 through
1991, the Joint Requirements
Oversight
Council*
validated
the requirements
for an advanced space-based
missile
warning
sensor to detect,
process,
and report
ballistic
missile
launches.
Air Force representatives
informed
us that the
associated
documents provided
guidance for the FEWS research
and
development.
FEWS
However, they stated that specific
requirements
contained
in a draft
October 1992 FEWS operational
requirements
document were never validated.
According
Secretary
options
capability,
and less
reducing
launched
idea that
to an October 1993 study3 performed
for the Under
for Acquisition
to review and recommend
for a future
space-based
infrared
surveillance
new needs could be met with a system that is simpler
costly
than FEWS. The study gave considerable
weight to
the size of the FEWS satellite
to allow it to be
on a smaller
vehicle
than Titan IV, which DSP uses--an
would reduce costs.
of Defense
The study stated
that although
there are strong reasons for DOD
wanting
a new, more able satellite
in the future,
(1) the current
requirement,
and associated
FEWS specification,
originated
in a
time of complex strategic
needs, (2) times have changed-'Conduct of the
April
1992.
*A group
Chairman
determine
analyses.
Persian
of high level
of the Joint
the validity
Gulf
War, Final
ReDOrt
to Congress,
DOD,
military
officers,
chaired
by the Vice
Chiefs of Staff,
having authority
to
of mission
needs and perform requirements
%pace-Based
IR Sensors,
October 1993, performed
by a technical
support
group from several
federally
funded research
and
development
centers
and referred
to as the Everett
study.
2
strategic
needs being less important
and global
awareness and
and (3) there is sufficient
theater
support
being more important,
time to review the requirements
and compete for a better,
simpler,
cheaper system within
the existing
budget constrained
The
study
recommended
that
the
requirements
be redone
schedule.
in context
of expected needs and capabilities
that could be
provided
by other ground- and sea-based systems.
DOD now plans to complete
a review of space-based
early
requirements
by October 1994.
Considering
the current
on tactical
needs, and presumably
a continued
need for
strategic
capability,
DOD must clearly
establish
a set
operational
requirements
that will
address these needs
making a substantial
investment
in a new system.
Cost-Effectiveness
warning
emphasis
a
of
before
Is Critical
Previous
studies
raised
questions
about the cost-effectiveness
of
FEWS and other advanced capabilities
relative
to DSP, and this
will
be a critical
matter
for ALARM. For example,
in 1991, we
reported"
that an Air Force cost and operational
effectiveness
analysis
showed life-cycle
costs for an enhanced DSP were
estimated
at $2.4 billion
to $3.5 billion
less than two
variations
of FEWS and a fully
capable Advanced Warning System.
We also reported
that a 1991 draft
study by a Defense Science
Board task force and a 1990 Air Force requirements
trade study
had similar
conclusions.
Also, part of the October 1993 study's
task was to identify
costeffective
options
for consideration
by DOD executives.
The study
presented
four options
that ranged in cost from $5.2 billion
to
$11 billion
for the period
2002 to 2015.
The lowest cost option
involved
down-sizing
the existing
DSP design and using mediumsized launch vehicles
instead
of the Titan
IV,
The highest
cost
option
involved
using a-lightweight
version
of FEWS, also
vehicle.
designed
for launch on a medium-sized
Cost and operational
effectiveness
analyses
are intended
to be a
management aid in decision-making
by illuminating
the relative
advantages
and disadvantages
of candidate
systems from among
plausible
alternatives.
Despite DOD's prior
experience
relative
to DSP, such an analysis
will
still
be critical
in justifying
a
new system.
Will
ALARM Be Affordable?
Based on Air
capable than
'Earlv
Force plans,
the
FEWS. To reduce
Warnincr Satellites:
Premature
(GAO/NSIAD-92-39,
initial
ALARM design will
near term costs,
the Air
Fundina for Follow-on
Nov. 7, 1991).
3
Svstem
be less
Force does
Is
not intend
to install
data processing
capabilities
and
communication
cross links
on the ALARM satellites--both
of which
were key features
in the FEWS design.
The Air Force claims this
helps address the affordability
problem that plagued FEWS.
Within
the fiscal
year 1995 Future Year Defense Program (FYDP)
the Air Force estimated
the costs for ALARM at $1.3 billion.
FEWS was estimated
at $4.6 billion--$3.3
billion
higher
than
ALARM for fiscal
years 1995 through
1999.
However, life-cycle
cost estimates
for both ALARM and FEWS, using
medium-sized
Atlas launch vehicles
are roughly
comparable.
ALARM
was estimated
at $11.3 billion
from fiscal
year 1995 through
2019.
Over the same period,
FEWS was estimated
at $11.7 billion,
or $400 million
higher than ALARM. In addition,
the Air Force
plans to upgrade ALARM starting
with satellite
number 5, but the
costs for these upgrades are not yet available.
than
While the cost estimates
show ALARM to be more affordable
FEWS in the short term, the total
life-cycle
costs lead us to
question
whether ALARM, with projected
upgrades,
will
actually
a more expensive
system.
Other
Factors
be
To Consider
Since program inception,
the Air Force has launched
16 DSP
satellites.
It currently
has multiyear
contracts
to procure
a
total
of 22.
To avoid a gap in coverage before making the
transition
from DSP to ALARM, the Air Force intends
to procure
one additional
DSP satellite
(number 23).
It is important
to
ensure that this procurement
will
provide
for an adequate
transition
and that any developmental
risk associated
with the
ALARM program is compatible
with such a transition.
In addition,
the launch vehicle
implications
for DSP 23 should be
considered.
DSP satellites
require
the heavy lift
capability
of
the Titan
IV launch vehicle,
which costs about $285 million
per
launch.
However, there is no Titan
IV available
for DSP 23 under
the current
contract
for 41 vehicles.
The Air Force and the
National
Aeronautical
and Space Administration
are discussing
the
possibility
of using the Space Shuttle
for a future
DSP launch.
MILSTAR:
A COSTLY AND
CONTROVERSIAL PROGRAM
During the past 12 years, DOD has invested
about $8 billion
in
the Milstar
program,
which has experienced
several
changes,
delays,
and cost increases.
On average,
each Milstar
satellite
placed in orbit will
cost about $1.3 billion--$1
billion
for the
satellite
and about $285 million
for the Titan
IV.
The first
satellite,
which was originally
scheduled
to be launched
in 1987
was actually
launched on February
7, 1994.
DOD expects to launch
the second Milstar
in May 1995.
4
Milstar
is designed
to be a highly
survivable
satellite
communications
system, particularly
resistant
to electronic
jamming,
for commanding and controlling
military
forces.
Originally,
the emphasis was on strategic
nuclear
warfighting
by
including
a low-data
rate communications
capability,5
primarily
for sending emergency action
messages to U.S. strategic
forces.
Tactical
forces were also planned users of this capability.
After
the fall
of the Berlin
Wall, congressional
leaders,
in
1990, considered
Milstar's
cost to be too high, its support
to
tactical
forces
inadequate,
and its nuclear
war-fighting
capabilities
unnecessary
for deterrence.
As a result,
the
National
Defense Authorization
Act for Fiscal
Year 1991 directed
the Secretary
of Defense to develop and carry out a plan for
either
a restructured
Milstar
or an alternative
advanced
communications
satellite
program.
DOD chose to restructure
the Milstar
program.
To lower costs,
it
decided to reduce the planned constellation
size from eight to
six satellites,
reduce the quantity
of other ground-based
equipment,
and eliminate
several
system survivability
features.
To provide
greater
utility
to tactical
forces,
it decided to add
a medium-data
rate capability
to satellite
4 and beyond.
In October 1992, based on guidance
from the conference
committee
on the fiscal
year 1993 Defense authorization
bill,
the Assistant
Secretary
of Defense for Command, Control,
Communications,
and
Intelligence
approved a further
reduction
in Milstar's
planned
constellation
size to four satellites.
At this point,
DOD's plan
was to launch the first
two satellites
based on the original
design with the low-data
rate capability.
Then, the medium-data
rate capability
would be added to satellite
3 and beyond.
In its October 1993 bottom-up
review of major defense programs,
DOD decided to keep Milstar's
constellation
size at four
satellites,
but to limit
the total
acquisition
to six satellites
--the
first
two, referred
to as Milstar
I, with the low-data
rate
capability
only,
and the next four,
referred
to as Milstar
II,
with both low- and medium-data
rate capabilities.
To reduce
long-term
costs,
DOD plans to replace
the Milstar
II design in
fiscal
year 2006 with an advanced capability
based on a smaller
satellite
design that will
use a smaller,
less expensive
launch
vehicle
than the Titan
IV.
'This low-data
rate capability
allows
information
to be
transmitted
at speeds ranging
from 75 to 2,400 bits per second
that would carry teletype
and compressed voice communications.
Medium-data
rate includes
speeds up to 1,544,OOO bits per second
that would carry regular
voice communications
and imagery.
5
Additional
Alternative
Cost Savinq
Should Be Assessed
we discussed
alternatives
for inserting
In our 1993 report,6
modern technology
into DOD's military
satellite
communications
programs that could reduce long term-costs
by about $17.6 billion
We specifically
discussed
an
compared with DOD's baseline
plan.
opportunity
for making a transition
to a common bus--a standard
satellite
platform
that supports
the mission
payload equipment,
which was one of several
DOD satellite
Regarding
Milstar,
communication
systems within
the plan, we suggested that such a
transition
could be made after
satellite
6. This was at a time
when DOD was planning
to build
eight Milstar
satellites,
thus the
acquisition
of satellites
7 and 8 could be avoided.
We
recommended that the Secretary
of Defense reassess various
alternatives
to preclude
the continuation
of costly,
customized
satellites.
In its December 1993 response to our report,
DOD (1) discussed
plans to terminate
Milstar
after
the sixth
satellite,
based on
(2) agreed with the need to move
the bottom-up
review decision,
unique busses toward common busses# and (3)
away from customized,
stated
that the most cost-effective
approach for inserting
modern
technology
was to begin developing
an advanced,
lower cost,
lower
weight payload
capability.
inserting
We now believe
there is a basis for DOD to consider
The
first
two
medium-data
modern technology
after
satellite
4.
rate Milstars
(satellites
3 and 4) are under development
and
However, a
scheduled
for launch in 1999 and 2000, respectively.
contract
has not been awarded for the last two Milstars
(satellites
5 and 6), which are planned to be launched in 2001
This would provide
a break point
in the
and 2002, respectively.
Milstar
program offering
an opportunity
to reduce costs through
technology
insertion.
Regarding
the insertion
of modern technology,
it was the
consensus of an outside
technical
support group, established
to
review options
and assess risk under DOD's bottom-up
review,
that
an advanced design could be deployed
as early as 2003 on a
This is in contrast
to DOD's
medium-sized
Atlas launch vehicle.
In fact,
some
planned deployment
of an advanced design in 2006.
members of the group believed
a first
launch of an advanced
satellite
would be possible
in 2000, using technology
already
developed
or under development.
"Militarv
Satellite
Communications:
of Dollars,
(GAO/NSIAD-93-216,
July
6
Opportunitv
9, 1993).
to Save Billions
If DOD did not acquire
satellites
5 and 6 and if it deployed
a
less-expensive,
advanced capability
in 2003, there would be a 2year delay,
from 2002 to 2004, in achieving
a four-satellite
constellation
with medium-data
rate capabilities.
DOD would have
to compare the benefits
of the potential
cost savings associated
with this approach,
which could be over $2 billion
including
launch costs,
to the operational
risk of not having a foursatellite
constellation
during the time period
now planned.
A
decision
would need to be made this year because the Air Force
for these satellites
in fiscal
plans to acquire
long lead items
year 1995.
In addition,
Milstar
satellites
require
the heavy lift
capability
of the Titan
IV launch vehicle.
This type of vehicle
costs about
$285 million
per launch.
There are no Titan
IVs available
for
Milstar
satellites
5 and 6 under the current
contract
for 41
vehicles;
therefore,
eliminating
these two satellites
would
reduce future
Titan IV procurement.
Mr. Chairman,
this concludes
my statement.
answer any questions
you or members of the
(707067)
I will
be happy to
Subcommittee
may have.
Ordering Information
The first copy of each GAO report and testimony is free.
Additional copies are $2 each. Orders should be sent to the
foIlowing address, accompanied by a check or money order
made out to the Superintendent
of Documents, when
necessary. Orders for 100 or more copies to be mailed to a
single address are discounted 25 percent.
Orders by ma&
U.S. General Accounting Office
P.O. Box 6015
Gaithersburg, KD 20884-6015
l
.
or visit:
Boom 1000
700 4th St. NW (corner of 4th and G Sts. NW)
U.S. General Accounting OEice
Wa&ington, DC
Orders may also be placed by &Iing (202) 512-6000
or by nsing fax number (301) 258-4066.
PRINTED ON
RECYCLED PAPER
united
states
General Accounting
Office
Washington, D.C. 20548
official
Penalty
Business
for Private
I
Use $300
Permit
No. GlOO